In retinopathy of prematurity (ROP), compromised physiologic vascularity and delay in physiologic retinal vascular development (PRVD) in Phase I precede vasoproliferative intravitreal neovascularization (IVNV) in Phase II. During prior grant periods, we found that intravitreal antibodies that inhibit VEGF-signaling reorder angiogenesis, which reduces IVNV and allows some PRVD, also impair pup growth, reduce physiologic vascularity and disorganize retinal structure in rodent oxygen-induced retinopathy (OIR) models. Targeting over-expressed VEGF or activated proteins in the VEGF-signaling pathway with cell-specific shRNA inhibits IVNV without reducing pup growth or physiologic vascularity but does not completely extend PRVD. In human infants with severe ROP we found that recurrent IVNV occurs due to incomplete PRVD and compromised physiologic vascularity after intravitreal anti-VEGF antibody. Together, these findings suggest that physiologic reorientation of angiogenesis to treat pathologic IVNV is insufficient to extend PRVD, but complete PRVD - with preserved physiologic vascularity - is important to prevent recurrent IVNV. Our next goal is to identify mechanisms to extend PRVD and preserve physiologic vascularity to prevent severe ROP by exploiting protective mechanisms endogenously or through exogenous treatment. Using models of maternal uteroplacental insufficiency (UPI) and pup OIR, we identified upregulated systemic erythropoietin (EPO) as a potentially-protective factor, which extended PRVD. We also found the protein, mediator of ERBB2-driven motility (MEMO1), localized to IVNV but not to normal developing vessels and interfered with EPO-induced mechanisms to inhibit retinal endothelial cell (EC) apoptosis and quench reactive oxygen species (ROS). MEMO1 was necessary for VEGF-induced angiogenesis mediated through the transcription factor, STAT3, in ECs. Using a lentiviral gene therapy method to specifically knockdown STAT3 in retinal ECs in vivo, we found EC-STAT3 knockdown inhibited IVNV in rat OIR. These findings support our goal for the next grant period to test the hypothesis that PRVD and physiologic vascularity can be modulated by EPO signaling through EPO- receptor (EPOR) in angiogenesis. Specific Aim 1 is to test if induced endogenous EPO signaling from UPI and pup-OIR is important for physiologic vascularity and regrowth by reducing ROS and apoptosis. Specific Aim 2 is to test if exogenous EPO is sufficient to promote PRVD and improve retinal structure and function without increasing IVNV in the setting of knocked down EC EPOR. Specific Aim 3 is to test if MEMO1 interferes with EPO-mediated mechanisms to promote PRVD, and if MEMO1 sustains VEGF-induced angiogenic signaling leading to IVNV. Methods include: novel lentiviral gene therapy in rat; oxygen-induced retinopathy models in rat and in novel composite murine dam UPI/pup OIR models; humanized knockin mice in which EPO signaling is hypoactive; Micron IV imaging; subretinal injections; optical coherence tomography; electroretinography; immunohistochemistry of flat mounts and sections; western blot, real-time PCR.